The Opportunity of Time-of-Use PricingFiled in:
Imagine if, instead of just my home shifting its AC, my entire community shifted its AC demand to off-peak hours, and what that could mean for the carbon intensity and electricity generation costs of my local grid.
One residential customer’s air conditioning experiment
Like many at RMI, I care about climate change and reducing the carbon intensity of the electric grid. I also care about my monthly electricity bill, and I’m keenly interested in keeping it as low as reasonably possible. If there are things I can do as a homeowner to tackle both topics, I’m all for it. And so, for example, I swapped out my home’s incandescent light bulbs for energy-efficient LEDs; they’ve reduced my monthly consumption and bill in one fell swoop.
But until earlier this year, there was a big elephant in the room—or, more accurately, mounted to the wall: our thermostat. It was old, with a broken screen so that we couldn’t read what setting it was on. My wife and I pressed the temperature up and down buttons until we got the house to what we thought was a comfortable temperature—without knowing exactly what the temperature was—and then simply pressed the “hold” button to keep our house at that temperature, 24 hours a day, 7 days a week. It’s a horribly inefficient (and expensive) way to condition a house, yet we muddled along that way for far too long.
WHEN AC AND PEAK GRID DEMAND COLLIDE
Finally, at the beginning of the summer we bought and installed a new programmable thermostat. I eagerly set up weekday and weekend programs designed to keep the house comfortable when we needed it to be, and operating in a low-energy mode when we didn’t. My wife is home with our young children during the day, so we don’t have the same opportunities as some folks to put their house into an energy-saving “away” mode during the work day, but we still did what we could.
Even during summer, where we live in Colorado gets nice and cool overnight, so we’d turn off the AC system and open the windows. In the morning when I woke up before work I’d close all the windows to lock in that cool air while the outside temperature climbed as the day progressed. The AC system, meanwhile, was programmed to turn on around noon, working hard through the hottest hours of the day to try and keep the inside of our house comfortable. Then the system would turn off later in the evening, after the sun had set and the outside temperature started to drop again.
That was all well and good except for one problem: such a schedule directly aligned my home’s AC use with the Colorado grid’s peak demand during the summer months. If you look at the grid’s load profile for residential customers like me during the months of July and August, demand really starts to climb around noon, peaking somewhere in the 6:00–8:00 p.m. range before tapering off to a low around 4:00–5:00 a.m. the next morning. By overlapping that load profile for the grid, the thermostat program I’d so enthusiastically set exacerbated peak demand and thus two issues I was trying to help fix: 1) By exacerbating grid peak I was helping force my local municipal utility to call upon less-efficient peaking plants (translation: more carbon emissions). 2) Those peaking plants are also pricier sources of generation, incurring added costs for my utility, and obliquely, eventually for me and other customers.
EXPERIMENTS IN PRE-COOLING
But then, inspired by a conversation I had with Jim Avery from California utility SDG&E, I decided to do something about it. I resolved to shift my AC demand, and set a new thermostat program to make it happen by precooling my house and minimizing my AC run time during the grid’s peak hours.
We still opened our windows at night and shut them in the morning. But now we ran the AC system during the mid to late morning, before grid demand started to climb and when our AC wouldn’t have to fight against the hottest hours of the day. During summer we like to keep our house around 75 degrees Fahrenheit, but now we precooled the house down to 70. When noon rolled around and grid demand was starting to climb significantly, however, our AC system was off. Instead of the system kicking on to work hard at cooling the house back down to 75 in the face of a 95-degree summer afternoon, we sat back and watched as—hour by hour—our home’s inside temperature slowly climbed back up to 75 from the pre-cooled 70. Later in the afternoon when our AC did have to turn on and do some work, it was cycling on and off, simply maintaining the house at 75, rather than running at full blast for hours on end during grid peak.
Normally I’d be tempted to celebrate this as a personal victory. We successfully shifted our AC demand to largely avoid system peak. But alas this story doesn’t have a completely happy ending. You see, aside from my personal motivations to make this happen, I should be financially incented by my retail electricity price structure to do things like this, but I’m not. My utility offers most residential customers a single rate option: an inclining block rate. I pay a per-kWh price for the first 750 kWh I consume each month, and a slightly higher price for the next block of kWhs above that. It doesn’t matter when I consume those kWhs, so the fact that I shifted my AC run time from grid peak to off-peak hours had zero impact on the price I paid for that electricity.
THE OPPORTUNITY OF TIME-OF-USE PRICING
As my RMI colleagues who wrote Rate Design for the Distribution Edge noted in their eLab paper, however, it’s time to evolve beyond the bundled, block, volumetric pricing utilities like mine continue to offer. We’re due to make a decisive move toward increasingly sophisticated electricity pricing, including time-of-use pricing that would financially incent customers like me to shift my energy management in ways that can benefit both me and the grid, like with precooling a home during off-peak hours.
It’s not such a big leap to offer customers a rate option that emphasizes that when you produce and consume electricity matters. Some utilities do offer optional time-of-use pricing programs, but such offerings are far from universal (witness my local utility and many others that don’t offer it) and far fewer customers are opting in to such time-of-use pricing than could be. More utilities should start offering a time-of-use pricing option for customers, and utilities can take another step and make time-of-use pricing adoption more widespread by making it the default structure with a chance for customers to opt out to a “regular” rate.
Such a time-of-use rate should be sufficiently granular as to make meaningful differentiations across times of day, and those blocks of time should correspond to real differences in cost of service based on demand during those times, not simply blocks of time for their own sake. In New York City, for example, utility ConEd offers electric vehicle owners a voluntary time-of-use rate that has off-peak and peak period pricing, with a “super peak” period embedded within the peak period (only for a restricted set of hours and certain months of the year). The peak period spans a whopping 16 hours of the day, from 8:00 a.m. through 12:00 a.m. If that truly reflects a higher cost of service because of elevated demand—New York is, after all, the city that never sleeps—great. But if there are other meaningful differences in cost of service within that 16-hour peak period, there may be further opportunities to offer an even more sophisticated time-of-use option for energy-savvy EV drivers and other customers.
Along similar lines, the price differences between the various times of day should be sufficiently large to truly influence customer behavior. The price signals I get for running my AC during off-peak or peak hours should more accurately reflect my cost of service during those times and be strong enough to motivate me to actually shift the hours during which I run the AC.
One of the country’s most progressive utilities moving forward with more sophisticated time-of-use pricing for customers is eLab member Sacramento Municipal Utility District (SMUD) in California. SMUD recently completed a two-year pilot study in which thousands of customers either opted in to or were defaulted in to time-of-use pricing structures that had up to four time blocks and prices—including on-peak and critical peak, similar to ConEd—based on cost of service.
The results of that pilot program were overwhelmingly positive. Both customers that opted in to and that were defaulted in to the time-of-use rates had significantly higher levels of overall satisfaction and higher levels of understanding of their utility bills than customers who stayed on their standard residential pricing. And whether customers voluntarily joined or were defaulted into the pilot, very few opted out of the time-of-use option.
Now looking ahead, time-of-use pricing will become the default pricing structure for all SMUD customers starting in 2018, though they’ll retain the option to opt out and go back to flat per-kWh pricing.
If we can do just that much for customers with more utilities across the country, imagine the humble possibilities. Imagine if, instead of just my home shifting its AC, my entire community shifted its AC demand to off-peak hours, and what that could mean for the carbon intensity and electricity generation costs of my local grid that still relies 75 percent on coal. I’ll of course continue to precool my home in future summers, but I hope the day is coming—and soon—when millions of residential customers like me will have a much clearer incentive, via pricing, to do that too.